skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: A Connection-Free Reliable Transport Protocol
The Internet Transport Protocol (ITP) is introduced as an alternative to the Transmission Control Protocol (TCP) for reliable end-to-end transport services in the IP Internet. The design of ITP is based on Walden’s early work on host- host protocols, and the use of receiver-driven Interests and manifests advocated in several information-centric networking architectures. The performance of ITP is compared against the performance of TCP using off-the-shelf implementations in the ns3 simulator. The results show that ITP is inherently better than TCP and that end-to-end connections are not needed to provide efficient and reliable data exchange in the IP Internet.  more » « less
Award ID(s):
1733884
PAR ID:
10405905
Author(s) / Creator(s):
;
Date Published:
Journal Name:
2020 IEEE 39th International Performance Computing and Communications Conference (IPCCC)
Page Range / eLocation ID:
1 to 6
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; (, 2020 16th International Conference on Network and Service Management (CNSM))
    The Internet Transport Protocol (ITP) is introduced to support reliable end-to-end transport services in the IP Internet without the need for end-to-end connections, changes to the Internet routing infrastructure, or modifications to name-resolution services. Results from simulation experiments show that ITP outperforms the Transmission Control Protocol (TCP) and the Named Data Networking (NDN) architecture, which requires replacing the Internet Protocol (IP). In addition, ITP allows transparent content caching while enforcing privacy. 
    more » « less
  2. ; (, ICN '20: Proceedings of the 7th ACM Conference on Information-Centric Networking)
    Named-Data Transport (NDT) is introduced to provide efficient content delivery by name over the existing IP Internet. NDT consists of the integration of three end-to-end architectural components: The first connection-free reliable transport protocol, the Named-Data Transport Protocol (NDTP); minor extensions to the Domain Name System (DNS) to include records containing manifests describing content; and transparent caches that track pending requests for content. NDT uses receiver-driven requests (Interests) to request content and NDT proxies that provide transparent caching of content while enforcing privacy. The performance of NDT, the Transmission Control Protocol (TCP), and Named-Data Networking (NDN) is compared using off-the-shelf implementations in the ns-3 simulator. The results demonstrate that NDT outperforms TCP and is as efficient as NDN, but without making any changes to the existing Internet routing infrastructure. 
    more » « less
  3. ; ; ; (, IEEE)
    Data centers require high-performance and efficient networking for fast and reliable communication between applications. TCP/IP-based networking still plays a dominant role in data center networking to support a wide range of Layer-4 and Layer-7 applications, such as middleboxes and cloud-based microservices. However, traditional kernel-based TCP/IP stacks face performance challenges due to overheads such as context switching, interrupts, and copying. We present Z-stack, a high-performance userspace TCP/IP stack with a zero-copy design. Utilizing DPDK's Poll Mode Driver, Z-stack bypasses the kernel and moves packets between the NIC and the protocol stack in userspace, eliminating the overhead associated with kernel-based processing. Z-stack em-ploys polling-based packet processing that improves performance under high loads, and eliminates receive livelocks compared to interrupt-driven packet processing. With its zero-copy socket design, Z-stack eliminates copies when moving data between the user application and the protocol stack, which further minimizes latency and improves throughput. In addition, Z-stack seamlessly integrates with shared memory processing within the node, eliminating duplicate protocol processing and serializationldese-rialization overheads for intra-node communication. Z-stack uses F-stack as the starting point which integrates the proven TCP/IP stack from FreeBSD, providing a versatile solution for a variety of cloud use cases and improving performance of data center networking. 
    more » « less
  4. ; ; (, IEEE/IFIP Network Operations and Management Symposium (NOMS 2023), May 2023)
    This paper presents mmCPTP, a cross-layer end-toend protocol for fast delivery of data over mmWave channels associated with emerging 5G services. Recent measurement studies of mmWave channels in urban micro cellular deployments show considerable fluctuation in received signal strength along with intermittent outages resulting from user mobility. This results in significant impairment of end-to-end data transfer throughput when regular TCP is used to transport data over such mmWave channels. To address this issue, we propose mmCPTP, a novel cross-layer end-to-end data transfer protocol that sets up a transport plug-in at or near the base station and uses feedback from the lower layer (RLC/MAC) to opportunistically pull data at the mobile client without the slow start and probing delays associated with TCP. The system model and end-to-end protocol architecture are described and compared with TCP and IndirectTCP (I-TCP) in terms of achievable data rate. The proposed mmCPTP protocol is evaluated using NS3 simulation for 5G NR (New Radio) considering a high-speed mobile user scenario. The system is further validated using a proof-of-concept prototype which emulates the high-speed mmWave/NR access link with traffic shaping over Gbps ethernet. Results show significant performance gains for mmCPTP over TCP and I-TCP (2.5x to 17.2x, depending on the version). 
    more » « less
  5. ; ; ; ; ; (, Arxiv)
    The volume of data generated and stored in contemporary global data centers is experiencing exponential growth. This rapid data growth necessitates efficient processing and anal- ysis to extract valuable business insights. In distributed data processing systems, data undergoes exchanges between the compute servers that contribute significantly to the total data processing duration in adequately large clusters, neces- sitating efficient data transport protocols. Traditionally, data transport frameworks such as JDBC and ODBC have used TCP/IP-over-Ethernet as their under- lying network protocol. Such frameworks require serializing the data into a single contiguous buffer before handing it off to the network card, primarily due to the requirement of contiguous data in TCP/IP. In OLAP use cases, this seri- alization process is costly for columnar data batches as it involves numerous memory copies that hurt data transport duration and overall data processing performance. We study the serialization overhead in the context of a widely-used columnar data format, Apache Arrow, and propose lever- aging RDMA to transport Arrow data over Infiniband in a zero-copy manner. We design and implement Thallus, an RDMA-based columnar data transport protocol for Apache Arrow based on the Thallium framework from the Mochi ecosystem, compare it with a purely Thallium RPC-based implementation, and show substantial performance improve- ments can be achieved by using RDMA for columnar data transport. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email